Computational Imaging With Dynamic Metasurfaces: A Recipe for Simple and Low-Cost Microwave Imaging

We review the evolution of metasurface antennas for computational microwave imaging and highlight the advantages and disadvantages of various configurations. As an illustrative example, we examine a computational imaging system comprising dynamic printed metasurface cavities as modular building blocks. These metasurface cavities can generate a multitude of spatially diverse, voltage-controlled illumination patterns, and can therefore encode a scene's reflectivity distribution into a set of measurements that can be postprocessed to produce an image. These single-port devices act as transmitters (Txs) and/or receivers (Rxs) and can be combined to create an electrically large aperture with a reduced number of (expensive and cumbersome) radio-frequency (RF) components. Here, we present some of the unique possibilities made possible by a dynamic metasurface imager. Specifically, we demonstrate high quality imaging with a reduced bandwidth of operation-as narrow as a single frequency. We then provide performance predictions for the case when dynamic metasurface cavities are used to image humans, as might be relevant in many security-screening applications. We conclude by describing the outlook of metasurfaces for imaging and identifying several future directions.